Chap. III. FALLING BANKS. 171 



fish, and so forth, has suggested itself to me. Their 

 rounded, water-worn appearance showed that they must 

 have been rolled about for a long time in the shallow 

 streams near the sources of the rivers at the feet of the 

 volcanoes, before they leapt the waterfalls and embarked 

 on the currents which lead direct for the Amazons. 

 They may have been originally cast on the land and 

 afterwards carried to the rivers by freshets ; in which 

 case the eggs and seeds of land insects and plants might 

 be accidentally introduced and safely enclosed with 

 particles of earth in their cavities. As the speed of the 

 current in the rainy season has been observed to be 

 from three to five miles an hour, they might travel an 

 immense distance before the eggs or seeds were 

 destroyed. I am ashamed to say that I neglected the 

 opportunity, whilst on the spot, of ascertaining whether 

 this was actually the case. The attention of Naturalists 

 has only lately been turned to the important subject of 

 occasional means of wide dissemination of species of 

 animals and plants. Unless such be shown to exist, it 

 is impossible to solve some of the most difficult problems 

 connected with the distribution of plants and animals. 

 Some species, with most limited powers of locomotion, 

 are found in opposite parts of the earth, without existing 

 in the intermediate regions ; unless it can be shown 

 that these may have migrated or been accidentally trans- 

 ported from one point to the other, we shall have to 

 come to the strange conclusion that the same species 

 had been created in two separate districts. 



Canoemen on the Upper Amazons live in constant 



